1 / 7

The Players

The Players. The LHC experiments To a lesser, but important extent, the EGEE apps SA1: most importantly, CERN-external sites JRA1 in various locations. Boundary Conditions. A finite amount of experienced manpower EGEE-1 ends March 2006

faris
Download Presentation

The Players

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. The Players • The LHC experiments • To a lesser, but important extent, the EGEE apps • SA1: most importantly, CERN-external sites • JRA1 in various locations ARDA Workshop, 21 October 2004 - 1

  2. Boundary Conditions • A finite amount of experienced manpower • EGEE-1 ends March 2006 • LCG-2 must be present on production infrastructure until something is proven better • Doesn’t exclude presence of e.g. gLite; doesn’t mean gLite must be integrated; only that gLite must not break the LCG-2 installation • No software can be deployed on large scale until a site-approved automatic install system works ARDA Workshop, 21 October 2004 - 2

  3. Customers • Many people believe they are the only customer of gLite • Even within HEP this is wrong • We have funding to do Grid via EGEE; like it or not this extends the # of customers • EVERYBODY forgets that SITES are customers of gLite. If a piece of sw ignores site constraints, it will not be deployed • R-GMA removal from EDG when backdoor was discovered • Nordugrid (or ARC) not deployed due to protocol mucking • Sites will • Support your jobs and data challenges (look at LCG-ROLLOUT) • Spend months of their time lobbying national funding agencies to buy machines to run your jobs, at very little cost to you • Have to clean up the mess when systems get hacked ARDA Workshop, 21 October 2004 - 3

  4. Problem and Proposed Solution HYPOTHESIS: Most of the heat is being generated since each of the PLAYERS listed above is being exposed to only a subset of the BOUNDARY CONDITIONS and therefore have a limited amount of sympathy for the other players. SOLUTION: Expose all parties to all constraints. POSSIBLE LOCATION FOR SOLUTION: the pre-production service: make this(and soon) the focal point of gLite activity. ARDA Workshop, 21 October 2004 - 4

  5. PPS Current Status • being installed now with LCG-2 core (see boundary condition 3) • eight committed sites (compatible with ARDA wish for more sites) • next-generation (Quattor) auto-install support will likely be used on at least one site, community support (SA1) is emerging; see boundary condition 4 ARDA Workshop, 21 October 2004 - 5

  6. What Needs to Change • streamline path from developers to PPS to as low as possible; get stuff there AFAP! • for the moment, no distributed tests on JRA1 testbed; just give us RPMs. • developers could give us stuff as raw as they like; however the speed at which things will be deployed of course will depend on how bad things are. • Again must not break LCG-2 service • ARDA project moves to PPS ARDA Workshop, 21 October 2004 - 6

  7. Process to Desired End Result • gLite development properly exposed to operational concerns and user concerns • gLite users properly exposed to development concerns and operational concerns • gLite resource providers properly exposed to (and able to influence) how system is deployed and used • as deployment matures, more and more sites can be brought in • at some point this becomes the production system with co-existing gLite and LCG-2 • LCG-2 is dropped when it can be. ARDA Workshop, 21 October 2004 - 7

More Related